search

Js2Hou / Subspace-Regression-based-Face-Recognition

We implement some sparse representation based face recognition algorithms here, including LRC, SRC, CRC, ESRC and etc.
1 stars 1 forks source link

SRC and ESRC cannot get output result #1

Open YMCLEARNER opened 2 years ago

YMCLEARNER commented 2 years ago

Hi! I am the one who currently using your Subspace-Regression-based-Face-Recognition as learning purpose. And when I try to run your code, all the classification algorithm can get accuracy properly except SRC and Euler SRC... When I try to run both of them, they keep running and won't get any output result (accuracy)... May I know how to let the SRC and Euler SRC algorithm run successfully in order to get the accuracy ? Thank you so much!

Image below shows that CRC and ECRC run properly but SRC keeps running:

1

Image below shows that LRC, RRC and ERRC run properly but ESRC keeps running:

image

Here is attached with my ipynb file which try to run all the algorithms:

Performance Evaluation.zip

Js2Hou commented 2 years ago

Wait a while. The SRC-based approach has no closed solution, which is solved by iteration. Therefore, it takes a long time.

YMCLEARNER commented 2 years ago

Wait a while. The SRC-based approach has no closed solution, which is solved by iteration. Therefore, it takes a long time.

Thanks for responding me! But I tried to let it run for more than 6 hours, it still won't get any output result (accuracy)... By the way, I tried to remove the # of print(prob.status) for checking that whether the SRC algorithm is running or not however it keeps output "optimal" and also won't get any accuracy result... May I know that how long that you wait for getting the accuracy of this SRC-based algorithm?

Image below shows that the time that I start for running the SRC algorithm:

1

Image below shows that the time that I run the SRC algorithm for more than 6 hours (it still don't get any accuracy result yet):

image